As advertising moves into conversational AI, brands are gaining a richer picture of consumer intent – not only what people search for, but what their intent behind their search is. In this guest article, Chris Mellish, MD at Accenture Song, Marketing Practice Lead UK and Ireland, argues that combining these new signals with retailers’ view of real-world purchasing behaviour could reshape how marketers operate.
OpenAI’s move into advertising has understandably set the industry alight – the select retailers who are advertising on the platform are seeing strong results. What makes this particularly interesting is not simply the arrival of another advertising channel, but the type of information that this new ‘conversational AI’ can bring into the purchase journey.
Traditional advertising has always chased signals: what people search for, click on, linger over, add to their basket. Conversational AI adds a whole different ball game – people don’t just tell AI what they want, they often explain why they want it. For example, a shopper might search ‘find me a carry-on for an upcoming San Sebastian trip’ – the AI can then start reasoning, and offering deals for a Spanish itinerary. This shift, from behavioural signal to stated reasoning will change how brands understand user intent altogether.
Our own data shows just how significant this shift could become – 71 percent of people expect generative AI to influence at least half of their spending over the next year, while 26 percent say AI has already led them to buy a more expensive product.
For brands, this opens a previously ‘locked’ door: not just knowing what someone is looking for, but understanding the reasoning behind the decision.
Retailers Still Have an Ace Card – What People Actually Bought
An AI assistant can become very good at understanding why somebody might choose a particular product, as it can understand the options, they are considering and the features that matter. What it cannot necessarily see is what happens when that intention meets reality.
People don’t always make purchasing decisions in the way they initially expected to as the product they intended to buy might be unavailable or they might see an alternative in store. This is where retailers have a direct line of sight. They have a direct view of what customers actually bought and over time, how those purchases develop into patterns of behaviour.
This is where retailers still hold a card AI doesn’t have: a direct line of sight into what people actually bought, and how that behaviour compounds into patterns over time. Conversational AI can illuminate motivation; retail data confirms outcome. The opportunity for brands is stitching those two together.
From Targeting Audiences to Shaping Consideration
This also reframes what advertising inside a chatbot actually does. Digital advertising has spent two decades getting sharper at finding the right audience and reaching them at the right moment. The next frontier may be less about reaching people and more about shaping what they consider in the first place.
When somebody asks an AI assistant which product they should buy, the important question for a brand is whether it is part of that conversation. Recommendations have always played an important role in helping people navigate choice, particularly when there are too many products, reviews and opinions to consider. What changes with conversational AI is the scale and immediacy of that recommendation.
That creates a different challenge for brands as being visible online is no longer necessarily enough. Brands need to be relevant to the context and trusted enough for the recommendation to influence the eventual decision.
People are more likely to trust technology when they understand how it is being used and feel that they have some control over the experience and the same will apply to AI-led recommendations. If consumers feel that a recommendation is simply an advert dressed up as advice, the value of the experience quickly diminishes.
Retail’s Role Gets Bigger, Not Smaller
It is understandable that retailers might look at the growth of conversational AI and see a potential threat to the value of their customer relationships.
As more of the consideration process moves into AI, the value of understanding what happens at the point of purchase arguably becomes greater and retailers are in a position to connect intent with outcome in a way that AI cannot necessarily do on its own, giving brands a way of grounding the new world of AI-generated intent in real customer behaviour.
Why Human Insight Matters More Than Ever
There is a temptation to think that as technology gets better at understanding people, we need to understand people less. The opposite is likely to be true.
As targeting and optimisation become increasingly automated, human understanding becomes more valuable. Marketers still need to understand motivations, but also the value exchange they are creating.
Technology is very good at delivering efficiency, relevance and frictionless experiences. But people don’t always make decisions in the most efficient way. We choose products because they feel familiar, because we trust the brand, because they fit how we see ourselves, or simply because something about the experience feels right.
That human understanding becomes particularly important when technology starts to play a more active role in the decision itself. If an AI assistant is helping someone narrow down their choices, brands need to think not only about whether they can be surfaced in that conversation, but why someone would feel comfortable acting on that recommendation.
And perhaps that is the more interesting implication of ChatGPT’s move into advertising. For brands, it’s not about creating another place to advertise, but about getting closer to the decision itself and understanding what makes someone move from consideration to action.
